Rune Hjorth

Aquatic Ecotoxicity Testing of Nanoparticles-The Quest To Disclose Nanoparticle Effects.

Abstract The number of products on the market containing engineered nanoparticles (ENPs) has increased significantly, and concerns have been raised regarding their ecotoxicological effects. Environmental safety assessments as well as relevant and reliable ecotoxicological data are required for the safe and sustainable use of ENPs. Although the number of publications on the ecotoxicological effects and uptake of ENPs is rapidly expanding, the applicability of the reported data for hazard assessment is questionable. A major knowledge gap is whether nanoparticle effects occur when test organisms are exposed to ENPs in aquatic test systems. Filling this gap is not straightforward, because of the broad range of ENPs and the different behavior of ENPs compared to “ordinary” (dissolved) chemicals in the ecotoxicity test systems. The risk of generating false negatives, and false positives, in the currently used tests is high, and in most cases difficult to assess. This Review outlines some of the pitfalls in the aquatic toxicity testing of ENPs which may lead to misinterpretation of test results. Response types are also proposed to reveal potential nanoparticle effects in the aquatic test organisms.

A critical analysis of the environmental dossiers from the OECD sponsorship programme for the testing of manufactured nanomaterials

In 2015, the OECD finally published the findings of its seven year testing programme for manufactured nanomaterials. Here, we present the first in-depth analysis of the published OECD dossiers with regards to data on physical and chemical properties, environmental fate and ecotoxicology. Each individual study in the dossiers was reviewed with regard to, among other, which OECD Test Guidelines (TG) were used, and the reliability assigned to the study. We furthermore analyzed in detail the suspension methods used, how media quality was quantified and physical and chemical characterization performed prior, during and/or at the end of the study. We find that the information in the dossiers present an incomplete portfolio of nanomaterial ecotoxicological evaluations that are difficult to draw substantive conclusions from and that most of the studies were not designed to investigate the validity of the OECD test guidelines. We acknowledge the effort of the OECD WPMN and recommend that a follow-on program is established with well-defined goals, end-points and direct funding to qualified research laboratories to ensure valid, rigorous, reproducible and efficient research.

What can nanosafety learn from drug development? The feasibility of “safety by design”

Abstract “Safety by design” (SbD) is an intuitively appealing concept that is on the rise within nanotoxicology and nanosafety research, as well as within nanotechnology research policy. It leans on principles established within drug discovery and development (DDD) and seeks to address safety early, as well as throughout product development. However, it remains unclear what the concept of SbD exactly entails for engineered nanomaterials (ENMs) or how it is envisioned to be implemented. Here, we review the concept as it is emerging in European research and compare its resemblance with the safety testing and assessment practices in DDD. From this comparison, it is clear that “safety” is not obtained through DDD, and that SbD should be considered a starting point rather than an end, meaning that products will still need to progress through thorough safety evaluations and regulation. We conclude that although risk reduction is clearly desirable, the way SbD is currently communicated tends to treat safety as an inherent material property and that this is fundamentally problematic as it represents a recasting and reduction of societal issues into technical problems. SbD therefore faces a multitude of challenges, from practical implementation to unrealistic stakeholder expectations.

Balancing scientific tensions

To the Editor — Within nanoecotoxicology, various scientific tensions exist in dynamic interplay. For example, exploratory and curiositydriven basic research competes with strategic-, appliedand regulatoryoriented approaches; research done within a paradigm of environmental chemistry must rival with a biological one; and research focused on well-controlled experimental set-ups vies with that prioritizing testing under complex and variable real-world conditions. All these research orientations are valuable, and just as in the concept of yin and yang, they should be regarded as complementary forces between which it is important to find a balance. Disconcertingly, however, the urgent need for environmental, health and safety data to support regulatory decision-making has created pressure to develop and use international standards in nanoecotoxicology, which can potentially cause an imbalance in the research orientations. Standardization may be necessary for reliability, comparability and regulatory relevance of nanoecotoxicological studies, but we should recognize and address the three double-binds that this creates. A ‘double-bind’ is a persistent type of dilemma whereby two choices are in tension, and success in one inevitably creates problems in the other. The first double-bind is a temporal one — it being arguably both too early and too late for standardization of nanoecotoxicology testing. Standardization is overdue because nano-products are already commercially available and data generated using existing methods may not be appropriate. It is too early, however, because results from initial testing are too diverse to offer any clear patterns on which the standardization of testing methods can be based. Even small modifications in sample preparation, composition of the test media or instrumentation can greatly influence the results of nanoparticle testing, making it difficult to condense current knowledge into ideal protocols. The second double-bind relates to the tension between pursuing tests under real environmental conditions and the need for well-controlled experimental set-ups. The Organisation for Economic Corporation and Development (OECD) spearheaded the development of testing guidelines for chemicals to ensure mutual cross-national acceptance of data that are relevant for classification, labelling and hazard assessment. However, this commitment to standardize experimental conditions tends to undermine the importance of testing under the complex and varied environmental conditions into which nanomaterials are released. The third double-bind concerns how pursuing knowledge according to any one particular approach inevitably creates selective ignorance, leading researchers to only partially understand complex phenomena. This does not imply that we are better off not pursuing knowledge at all, but rather highlights the importance of being aware of how paradigms shape and delimit what knowledge is produced. Given the complexity of nanoparticle properties, experimental media and other influencing factors, at present there is no single paradigm or testing approach that can be claimed to outrank others. Under such high uncertainty, it is important to actively reflect on the assumptions and limitations for all approaches. The triple double-bind situation facing nanoecotoxicology standardization can seem paralysing. Although there may be no ultimate escape, one way to move forward in a responsible manner is to acknowledge, actively cultivate and support the diverse research areas in nanoecotoxicology as an essential condition for scientific robustness. Funding bodies, researchers, regulators, publishers and standardization organizations alike need to acknowledge the triple double-bind situation. Funders should ensure that there are opportunities for diverse research approaches to be pursued. Researchers should continue to trial, develop and report on different test methods and communicate the limitations of their knowledge transparently. Scientific journals should remain open towards publishing different types of research (even when consensus seems to be emerging in favour of standardized testing), as well as publishing negative results. Regulators should resist the assumption that public authority requires scientific certainty and should be nuanced in their evaluation of data quality — focusing on both whether a study follows a particular standard and whether it is properly described, well performed and uses well-characterized nanoparticles. Finally, standardization organizations should resist the urge to develop premature standards and remain open, flexible and responsive to ongoing exploratory research and testing. It is now clear that standards for chemical testing are not appropriate for nanoparticles. However, rushing to develop harmonized standards for nanoecotoxicology and ignoring the three double-binds we face may unwittingly cause undue imbalance in the field. This could create false negatives that would place human and environmental health and safety in danger, and threaten the scientific and public legitimacy of nanoecotoxicology research. ❐

The role of alternative testing strategies in environmental risk assessment of engineered nanomaterials

Within toxicology there is a pressure to find new test systems and organisms to replace, reduce and refine animal testing. In nanoecotoxicology the need for alternative testing strategies (ATS) is further emphasized as the validity of tests and risk assessment practices developed for dissolved chemicals are challenged. Nonetheless, standardized whole organism animal testing is still considered the gold standard for environmental risk assessment. Advancing risk analysis of engineered nanomaterials (ENMs) through ATS was discussed in September 2014 at an international Society for Risk Analysis (SRA) workshop in Washington, D.C. and serves as the point of departure for this paper. Here we present the main outcomes by describing and defining the use of ATS for ENMs as well as discussing its future role in environmental risk science. We conclude that diversity in testing should be encouraged to avoid “selective ignorance” and that, through an iterative process with low-tier and high-tier testing, data-generation can be validated to ensure relevant endpoints. Furthermore, simplified screening of ENMs could enable early decision-making on material design, while complex multispecies studies should be utilized to skip uncertain environmental extrapolations and give rise to more accurate risk analysis.

The shortfall of risk assessment for decision-making

To the Editor — In August 2017, Nature Nanotechnology published a focus issue on the role of environmental risk assessment (RA) and life-cycle assessment (LCA). The focus was on the challenges of combining these two tools to get a better and more comprehensive assessment of the impact of engineered nanomaterials (ENMs)1–3. Importantly, the roles of RA and LCA should not be separated from a larger discussion on how their output feeds into decision-making and enables responsible risk management. It seems almost implicit that we expect such tools to be able to guide regulatory decisions, industrial design choices and our acceptance of risk. But how to assess the risk, and therefore the safety, of ENMs is still very much an ongoing scientific research topic4,5. For years, there has been a consensus in the scientific community that we are not able to provide the information needed for quantitative evaluation of the environmental effects of ENMs, and, according to a recent review by Hjorth et al.6, this is unfortunately still the case today. This is despite a substantial number of ecotoxicity studies that have been published over the past few years. The overarching issue is that the generated data generally suffer from low reproducibility and reliability, owing to a convoluted and time-dependent exposure during testing. This limits meaningful comparison between studies and — in the end — the usefulness of ecotoxicity studies for RA and LCA. We are, furthermore, lacking reliable exposure data as environmental concentrations are predicted with only limited knowledge on the production, release and transport of ENMs, and without the possibility of analytical validation7. Although there is a disagreement in the literature on the general applicability of the RA concept for ENMs, it should be uncontroversial to argue that RA cannot currently be used to accurately assess the environmental risk posed by ENMs, the main roadblock being lack of reliable input data, especially long-term toxicity data6 and as a close second, lack of exposure data and analytical measurements7. In other words, the primary issue with assessing risk is not the lack of adequate quantitative tools — it is the data needed to feed them. Intuitively, when quantitative assessment of risk falls short, we should turn to more qualitative assessments. Back in 2012, Grieger et al.8 evaluated various alternative risk analysis frameworks and concluded that they generally had the ability to incorporate qualitative elements as well as to account for uncertainty and LCA perspectives, but also that most had an occupational health focus, few considered environmental risk, and almost all frameworks needed further testing. Disconcertingly, the European Commission’s FP7 project MARINA could recently only repeat the same conclusion that qualitative RA tools still need testing and validation, while stressing that the field of nanosafety is years from being able to apply more quantitative assessments9. This does not mean that estimating risk cannot be pursued through RA, but it highlights the inadequacy in relying on the output of such analysis in solitude and the pressing issue of enabling decision-making through other means. A good example of this is the biocidal use of nanoscale silver, which today is the most used ENM in consumer products and also served as the main case in the focus issue1. The potential side-effects of nanoscale silver have been heavily scrutinized, but any conclusion on whether it poses a human and environmental risk is effectively stuck in evaluation10, and its biocidal use in Europe is currently awaiting further risk assessment not due until the end of 2024 (ref. 11). However, RA is unlikely to solve this issue. A noticeable parallel to nanoscale silver is the biocidal use of triclosan; and although triclosan has been used in consumer products for 50 years, we are still uncertain to what degree this is an issue for human and environmental health. For instance, just as for nanoscale silver, the risk of antimicrobial resistance remains unresolved, which is striking given that no other biocide has been investigated as thoroughly12. Recently, the US Food and Drug Administration (FDA) decided to remove triclosan and 18 other active ingredients in antibacterial soaps13, starting in September 2017. The FDA had notably not demonstrated a quantitative or qualitative evidence of harm; instead the chemicals are unwanted as they fail to add benefit to the products that would warrant any potential risk. A similar disapproval of the use of triclosan without evident benefits has also been expressed by the Scientific Committee on Consumer Safety in Europe12. Looking at risk with blinkers, without weighing the accompanying benefits, is an inherent limitation of RA, and leaving out this context makes it arguably harder to deem a risk acceptable or not. With the level of uncertainty present in the risk evaluation of ENMs, we need to be able to justify taking the risk in the first place. Looking at the function and benefit of a material is also the first step of conducting an alternatives assessment, which is increasingly being incorporated into corporate and legislative practice, and utilization of such comparative tools could provide the framework to help enabling decision-making on the best and safest use of ENMs14. So, despite the continuous improvement in many important areas related to the environmental risk of ENMs (for example ecotoxicity testing and exposure modelling), we must come to terms with the widespread uncertainty present in RA of ENMs and acknowledge that it is likely to persist. Decision-making is required in spite of scientific uncertainty15. With few risk management options available, a more holistic risk–benefit approach can enable decision-making and help to limit exposure, especially to imprudent biocidal use of ENMs. Although we should temper our expectations regarding what can currently be achieved through RA of ENMs, this should not stop us from making rational decisions on its use. ❐